Micro-RNAs and aging

Francisco J. Enguita UBCe, IMM [email protected]

The basics

Aging theories (some ...) Molecular Theories 1. Codon restriction - Fidelity/accuracy of mRNA translation is impaired due to inability to decode codons in mRNA. 2. Somatic mutation - Accumulation of molecular damage, primarily to DNA/genetic material. 3. Dysdifferentiation - Gradual accumulation of random molecular damage impairs regulation of gene expression. 4. Gene regulation - Aging caused by changes in gene expression regulating both aging and development.

Cellular Theories 1. Wear and tear - accumulation of normal injury (weak theory). 2. Free radicals - Oxidative metabolism produces highly reactive free radicals that subsequently damage protein and DNA. 3. Apoptosis - Programmed cell death resulting from genetically determined events or genome crisis. 4. Senescence - Phenotypes of aging are caused by an increase in frequency of senescent cells. Senescence may be the result of telomere loss (replicative senescence) or cell stress (cellular senescence).

The AGING DOGMA(s): 1. It is difficult to determine cause from effect in aging theories, many theories are based on an observation of some parameter that changes with age. However, it is difficult to determine if a change in function is a cause or an effect of the aging process. 2. We do not know what causes aging, a combination of theories may be correct, or some theories may be correct only in specific organisms. 3. Models to study aging are often inaccurate and difficult to set up.

LIFESPAN (years)

Cell aging (Cellular damage) Molecular damage (DNA, RNA, Protein, Carbohydrates, Lipids)

• Reactive oxygen Species (ROS) • Free radicals

• Cellular metabolism

• Errors in biological Processes (DNA replication..)

• UV-light • Chemicals • External inducers

Epigenetics landscape How a cell, without changing its genome, can undergo differentiation along highly diverse pathways?. The current vision of the uses the movement of pinballs to illustrate the complex and bidirectional epigenetic control of cell development and differentiation. The movement (representing different developmental stages) of the ball (the cell) in the machine depends on many epigenetic effectors (seen as flippers, obstacles, etc.) including DNA methyltransferases; histone modifiers, chromatin-remodeling factors and ncRNAs

Kalun & Fraga, J. Gerontol., 2009

Proposed role for miRNA regulation vs age

Postulated miRNA role vs cell aging signals

Bates et al, 2009

Postulated miRNA role vs cell aging signals

Chen et al, Ageing Res. Rev., 2010

Experimental design and results

Our cell model Skin fibroblasts

Young individuals (3, 10, and 11 y/o)

Old individuals (73, 81 and 87 y/o)

Low density RT-PCR

miRNA-RT array cDNA synthesis

Cell or tissue sample TRIZOL extraction

Total RNA Small RNA (< 200 nt) Real-time PCR

Young group

miRNA expression changes in the cellular aging model

Old group Dot‐plot diagram showing the differences in expression of 384 human miRNAs between the  analyzed two sample groups. The magenta line corresponds to 4 fold change in expression. 

miRNA expression changes in the cellular aging model Micro-RNA

Fold-change (old vs young)

P-value

hsa-miR-146a

12.77

0.056

hsa-miR-200c

3.07

0.033

hsa-miR-302c

6.25

0.013

hsa-miR-196a

-12.46

0.039

hsa-miR-134a

2.54

0.056

hsa-miR-218

2.10

0.015

hsa-miR-335

8.43

0.034

hsa-miR-376a

2.20

0.033

hsa-miR-139-5p

2.85

0.038

hsa-miR-600

3.23

0.058

hsa-miR-567

6.50

0.048

hsa-miR-208

6.60

0.030

Quantitative  results  of  selected  miRNAs  after  profiling  by  RT‐ PCR  arrays  and  normalization  against  U6  RNA.  Experimental  samples were structured in two  groups,  young  cells  (fibroblasts  isolated  from  human  individuals  with  3,  10  and  11  years  old),  and  old  cells  (fibroblasts  isolated  from  human  individuals  with  73,  81  and  87  years  old),  and  the  results  compared.  The  table  shows  the  statistically  significant  values  for  miRNAs  with  differential  expression  between  aged  and  young  cells  and  the  corresponding  fold  changes and p‐values. 

miRNA expression changes in the cellular aging model

Young 

Old 

Young 

Old 

Pathways and networks analysis

Overexpressed aging miRNAs: putative target analysis Predicted targets Experimentally validated targets

600 500

SOURCE: Predicted : Targetscan Validated: MirWalk

400 300 200 100 0

m iR -1 46 a m iR -2 00 c m iR -3 02 c m iR -1 34 m iR -2 18 m iR -3 35 m iR -3 76 m a iR -1 39 -5 p m iR -6 00 m iR -5 67

Number of targeted transcripts

700

Overexpressed aging miRNAs: targets

Genes from GENAGE database

Validated targets for the agingmiRNAs

35 261 479

Aging miRs: network analysis % genes in path Impact factor Wnt signaling pathway VEGF signaling pathway Thyroid cancer TGF-beta signaling pathway Small cell lung cancer Renal cell carcinoma Prostate cancer Pancreatic cancer p53 signaling pathway Non-small cell lung cancer mTOR signaling pathway Melanoma MAPK signaling pathway Jak-STAT signaling pathway Glioma Focal adhesion ErbB signaling pathway Cytokine-cytokine interaction Colorectal cancer Chronic myeloid leukemia Cell cycle Bladder cancer Apoptosis Adipocytokine signaling pathway Adherens junction 0

5

10

15

20

25

30

Networking: Pathway Express

35

Aging miRs: pathway analysis

Aging miRs: pathway analysis

Aging miRs: pathway analysis

Aging miRs: network analysis BIOCARTA based analysis

Networking: ClueGO Processing: Cytoscape

Aging miRs: network analysis KEGG based analysis

Networking: ClueGO Processing: Cytoscape

Aging miRs: recent developments

Aging miRs: aged-related protein networks

Regulatory position of some of the upregulated miRNAs in aged cells within the hubs of agerelated protein networks (agerelated networks is a concept recently discused by Wolfson et al, 2009, Int. J. Biochem. Cell Biol., 41, 516-520). Targetscan program was used to predict the targets for the overexpressed miRNAs. Arrows from the miRNA names indicate putative post-transcriptional control of the selected miRNA over the corresponding mRNA.

Proof-of-concept : miR200c – AKT1

Aging miRs: proof-of-concept AKT1 vs mir-200c Gene

microRNA

Position

Seed

dGduplex

dGopen ddG

AKT1_3UTR

mir200c

456

6:1:1

-14.5

-7.03

-7.46

AKT1_3UTR

mir200c

923

6:0:0

-13.8

-6.72

-7.07

AKT1_3UTR

mir200c

632

6:1:0

-7.99

-6.18

-1.80

AKT1_3UTR

mir200c

201

6:1:1

-8.5

-7.62

-0.87

AKT1_3UTR

mir200c

625

6:1:1

-9.01

-8.50

-0.50

AKT1_3UTR

mir200c

276

6:1:1

-8.2

-8.41

0.21

AKT1_3UTR

mir200c

64

7:1:0

-10.4

-10.79

0.39

AKT1_3UTR

mir200c

75

6:1:1

-6

-6.40

0.40

AKT1_3UTR

mir200c

309

6:1:1

-6.7

-9.35

2.65

AKT1_3UTR

mir200c

661

6:1:1

-8.8

-12.70

3.90

AKT1_3UTR

mir200c

932

6:1:0

-0.54

-6.48

5.94

AKT1_3UTR

mir200c

584

7:1:1

-8.52

-15.36

6.84

Aging miRs: proof-of-concept AKT1 vs mir-200c Young Group

miR-200c expression levels

Old Group

AKT1

13 12

Tubulin

11 10 9

2^-Delta Ct

8 7 6 5 4 3 2 1 0

3 yrs

10 yrs 11 yrs 73 yrs 81 yrs 87 yrs

Sample groups

Aging miRs: proof-of-concept AKT1 vs mir-200c

Relative Luciferase Activity

1.0

Empty vector No RNA Scramble miR-200c

0.8

0.6

0.4

0.2

0.0

AKT1 Tubulin

Transcriptomic analysis

Aging miRs: open questions 1. Role of aging-miRs in gene expression control 2. Mechanisms of putative regulatory effects 3. Role of alternative splicing events in the regulatory mechanism (poly-A ?).

Models used in this study Affymetrix GeneChip Human Exon 1.0 ST Array

AGING (Localdata) Human fibroblasts (3,10 and 11 yo) 3 arrays

Human fibroblasts (73,81 and 87 yo) 3 arrays

PROGERIA (Cao et al, J. Clin. Invest., 2011) Human fibroblasts 6 arrays

Human HGPS fibroblasts 4 arrays

SENESCENCE (Cao et al, J. Clin. Invest., 2011) Human fibroblasts 12 arrays

Human fibroblasts (replicative senescence) 12 arrays

Splicing events in aging, senescence and Progeria cells

Inclusion

Exclusion

Alternative splicing events

Splicing events: comparison Number of AS events 3 events in genes within GENage

6 events in genes within GENage

0 events in genes within GENage

Gene expression analysis Upregulated transcripts

Downregulated transcripts

6%

Gene expression analysis in aging cell model and targets for aging-miRs 1% Upregulated transcripts

Upregulated transcripts Total 690

Predicted targets of aging miRs Validated targets of aging miRs

92% 3% 1%

Downregulated transcripts

Downregulated transcripts Total 617

95%

Predicted targets of aging miRs

Models for miRNA global regulation of gene expression Translational repression Decrease of mRNA stability miRNA

target transcript

?

Increase in miRNA levels

Increase in target transcript levels

target transcript miRNA

target transcript

Translational repression Decrease of mRNA stability (Inverse correlation)

Conclusions 1. Upregulated miRNAs are predominant in our cell model of human aging (Aging-miRs). 2. Target analysis for the upregulated miRNAs showed genes involved in pathways already described and characterized in several aging models (Wnt signalling, insulin signalling, TGFbeta signalling, ErB signalling, etc). 3. Network analysis allowed to determine the putative control of the selected miRNAs over transcripts within aging networks. 4. AKT1 is a target of miR-200c. 5. Expression levels of Aging-miRs are not inversely correlated with the levels of their corresponding predicted and validated targets. 6. AS events in aging apparently do not have a strong influence over the differential regulation by miRNAs in aged and non-aged cells.

AKNOWLEDGEMENTS • André Melo (IMM, Lisbon, PT) • Javier Martínez (IMBA, Viena, AT) • Stefan Erkeland (Erasmus University Medical Center, Rotterdam, NL) • Marina Costa (IMM, Lisbon, PT) • Maria Carmo-Fonseca (IMM, Lisbon, PT)